camels from roman imperial sites in serbia

281ANTHROPOZOOLOGICA • 2014 • 49 (2) © Publications Scientifiques du Muséum national d’Histoire naturelle, Paris.

Key words Camels,

Bactrians, camel hybrids,

Roman archaeozoology,Viminacium,

Sirmium,Vranj,

amphitheatre,villa rustica.

Sonja VukoVić-BogdanoVićLaboratory of Bioarchaeology, Department of Archaeology,

Faculty of Philosophy - University of Belgrade, Cika Ljubina 18-20, Belgrade (Serbia)

[email protected]

Svetlana BlažićMuseum of Vojvodina Dunavska,

35 Novi Sad (Serbia)[email protected]

Camels from Roman imperial sites in Serbia

Vuković-Bogdanović S. & Blažić S. 2014 — Camels from Roman Imperial sites in Serbia. Anthropozoologica 49 (2): 281-295. http://dx.doi.org/10.5252/az2014n2a09.

AbstrActThis paper presents camel remains identified in Roman sites in Serbia. The remains originate from Viminacium, the capital of Roman province Moesia, and from the 3rd-4th century villa rustica in the locality of Vranj, in the prov-ince of Pannonia. In Viminacium, they were found in a 4th century dump, near the Eastern necropolis of the city, and in the 4th century layers in the area of the amphitheatre. In order to conclude whether these bones belonged to Camelus dromedarius or Camelus bactrianus, measurements and morphology of our specimens were compared with camel bones found at other Roman sites, and with modern specimens. We have identified two-humped camels and also hybrid individuals. The role of the camels is also discussed here: whether they were used in public games (ludi), for military purposes, or as transportation animals. As a conclusion, we acknowledge that camels were not rare animals in Roman provinces as it was considered before.

282 ANTHROPOZOOLOGICA • 2014 • 49 (2)

Vuković-Bogdanović S. & Blažić S.

IntRoduCtIon

one-humped camels, Camelus dromedarius, nowa-days occupy arid areas from north Africa to the Caspian sea and from Arabia to northwestern India, while two-humped camels, Camelus bactri-anus, spread from the Caspian sea to Central Asia (Zeuner 1967).

unlike on other common domesticates, data that deal with camel domestication are very scarce. It is considered that dromedaries and Bactrians are domesticated independently from different wild sub-species of Camelus ferus (Peters & driesch 1997). The fact that there are no clear morphologi-cal differences between wild and domestic camels complicates the understanding of domestication (olsen 1988). Ancient writers were familiar with both camel species. Aristotle (Hist. Anim. B. II. c. 1) wrote that ‘The Bactrian camel differs from the Arabian in having two humps as against the latter’s one’, while Pliny (Plin. nat. 8.26) took this ac-count from him.

According to camel skeletal remains that were found at the site of Shar-I Sokhta in Iran, it was considered that Bactrian camel was domesticated in the region ca 4500 years ago (Zeder et al. 2006 after Compagnoni & tossi 1978). More recent ar-chaeozoological evidences moved the center of camel

domestication to the east. Camelid skeletal remains, which were found in neolithic sites (6100-5300 BCE) in Inner Mongolia and China (olsen 1988) are considered to be of domestic Bactrian (Peters & driesch 1997). By the middle of the 4th millennium Bactrian camels were already in turkmenistan and by the middle of the 3rd millennium B.C., they were present in Bactria and the surrounding area to the north and south (Potts 2003).

Evidence of early domesticated dromedaries comes from the sites in the area of Arabian Peninsula (Umm an-Nar, etc) and date back to the 1st half of the 3rd millennia BC. After it was domesticated, one-humped camel started to spread to the east in the beginning of the 2nd millennium BC and until the middle of the millennium it reached its most easterly point -India (Köller-Rollefson 1996). towards the end of the 2nd millennia BC, camels started to move with camel riding tribes to the center of the Middle East (Köller-Rollefson 1993: 184). It is uncertain when the one-humped camel reached north Africa, but it is believed that it was introduced either in the Iron Age or by Romans, where Septimius Severus (194-211 Ad) moved Syrian troops to fight nomads (Morales Muñiz et al. 1995).

Both camel species appeared in Europe in the Ro-man period and in the Middle Ages. It is believed that Bactrian camel reached Europe first and then

Mots clésChameau,hybrides,

Archaeozoologie romaine,Viminacium,

Sirmium,Vranj,

amphithéâtre,villa rustica.

résuMéDes chameaux de site impériaux de la Serbie.Cet article présente des restes de chameaux découverts sur des sites archéolo-giques romains de Serbie. Les restes proviennent de Viminacium, capitale de la province romaine de Moesia et d’une villa rurale du 3e-4e siècle, dans la localité de Vranj, dans la province de Pannonia. à Viminacium, ils ont été trouvés dans les dépôts du 4e siècle près de la nécropole orientale de la cité et dans les niveaux du 4e siècle dans la zone de l'amphithéâtre. Pour décider si ces os appartenaient à Camelus dromedarius ou Camelus bactrianus, les mesures et la morphologie de nos spécimens ont été comparées avec celles des ossements de chameaux d’autres sites romains. nous avons pu déterminer des chameaux à deux bosses et des hybrides. Le rôle du chameau est aussi discuté ici : utilisation pour les jeux publiques (ludi), à des fins militaires ou comme bête de somme. nous concluons que les chameaux n’étaient pas si rares qu’on l’aurait cru auparavant dans les provinces de l’empire romain.

283

Camels from Roman Imperial Sites in Serbia

ANTHROPOZOOLOGICA • 2014 • 49 (2)

turned up in southern Russia between the 5th and the 3rd century BC. Rare finds of Bactrians were also found among animal bones from Greek colonies on the northern coast of the Black sea (Bökönyi 1974: 226-227).

Camel remains are known from Roman period sites in Italy (Mazzorin 2006), the Iberian Penin-sula (Morales Muñiz et al. 1995), Austria (Riedel 1999), Switzerland, ukraine (Bökönyi 1974), France (Clutton-Brock 1987), Germany (Benecke 1994), England (Applebaum 2002), Slovenia (Bartosie-wicz & dirjec 2001; Bartosiewicz 1999), Hungary (Bartosiewicz 1995; Bartosiewicz 1996; Bökönyi 1989; Bökönyi 1974) and Bulgaria (Beech 2007; Schramm 1975).

In Serbia1 camel bones were found in three Roman sites: Sirmium, Viminacium and Vranj. Sirmium camel finds are known from R. Lauwerier’s (1978) unpublished thesis. In this paper we are studying the morphology and osteometry of camel remains from Viminacium and Vranj, along with the usage and cultural implications of those animals in the Roman world, and particularly on the territories of the Roman provinces of Moesia and Pannonia.

ARCHAEoLoGICAL And ARCHAEoZooLoGICAL BACKGRound

Viminacium

Viminacium is situated on the right bank of the river Mlava (Fig. 1), close to its confluence with the danube, in today's Eastern Serbia. It was founded as a military camp where Legion VII Claudia was stationed from the beginning of the 1st century. By the camp, a city arose, which became the capital of province of Moesia Superior and later Moesia Prima. Archaeological excavations of Viminacium, conducted by the Institute of Archaeology in Belgrade, since the seventies to the present day, have mainly been fo-cused on the necropolis (Korać & Golubović 2009).

1. Since this paper was submitted for publication more camel bones were unearthed in Roman sites in Serbia. Among them is an almost entire camel skeleton from Viminacium (Vuković & Bogdanović 2013) and a few specimens from two other sites in Serbia: Davidovac-Gradište, near Vranje and Pirot, which have not been published yet.

Fig. 1. — Map of the provinces of Moesia Prima and Pannonia Secunda in the 4th century AD with marked archaeological sites in Serbia where camel bones were discovered.

In the city and the camp, only the thermae and north-ern entrance gate of the camp have been excavated so far, while the systematic excavations of the city amphitheatre have begun in 2007 and are still under way. The amphitheatre was built in the first quarter of the 2nd century Ad and it was used until the begin-ning of the 4th century Ad (nikolić & Bogdanović 2012). Analyses of faunal material from Viminacium amphitheatre are also in progress and until now, 15 camel bones were found. one camel bone was also found in the dump of the Eastern necropolis. Fauna from the Roman city of Viminacium (Vuković 2010) consists mainly of domestic animals: cattle, pigs, equids, dogs, sheep, goats, cats and camels, while wild animals: bears, wild boars, red deer and hare are also present although in small numbers.

Sirmium

Sirmium is located on the left bank of the river Sava (Fig.1), beneath the modern city of Sremska Mitrovica, in present day northern Serbia. The city was the capital of the province of Pannonia Secunda in late antiquity. After the establishment of tetrarchy, Sirmium was one of four capitals of the Roman Empire. Archaeological excavations of Sirmium, conducted by the Institute of Archaeology in Belgrade and Museum of Srem have been carried



out continuously since the mid 20th century. The most important excavated objects include: imperial palace, hippodrome, horrea, Licinus thermae and Jupiter sanctuary (Milošević 2001). The fauna of Sirmium (Gilić 1994; nedeljković 1997; nedeljković 2009) consists mainly of domestic animals (cattle, pig, sheep, goat, horse, dog, cat, poultry), while wild animals (aurochs, red deer, row deer, hare, wolf, wild birds) represent less than 5% of the total assemblage. In the faunal material from the ‘site 45’, which is known from R. Lauwerier (1978) unpublished thesis, 11 camel bones were found. Among them two phalanges dated back to the 4th century Ad, while seven articulating vertebras and one scapula and calcaneus were from mixed Roman and Mi-gration period layers (Lauwerier 1978). Sirmium camel bones were not at our disposal, so we have not included them in this study.

Vranj

Locality of ‘Vranj’ is situated in the village of Hrt-kovci (Fig.1), close to the left bank of the Sava river, in present-day northern Serbia. In the period between 1991 and 2004 in Vranj, a Roman villa rustica was excavated by the Museum of Vojvodina. The villa, which was located in the wider region of

Sirmium, was an agricultural establishment with well equipped rooms. The villa was continuously habitated from the mid 3rd century, until the mid 4th century (Ruševljan 2004; Ruševljan 2005). Among faunal remains from Vranj (Blažić 1993) domestic animals: cattle, sheep, goat, pig, horse and dog predominate, while wild animals: red deer, roe deer, wild pig and wolf, are present in small number. In this assemblage three camel bones were found.

MAtERIALS And MEtHod

We have analyzed sixteen camel bones from Vimi-nacium and three camel bones from Vranj. Vimina-cium camel finds include: atlas, three thoracic and one lumbar vertebra, one rib, distal fused radius and ulna, a distal femur, distal tibia, an astragalus and two first phalanges. Majority of those specimens were found in the area of the amphitheatre, while one phalanx was found in the dump of the Eastern necropolis in locality of Pirivoj (Vuković 2010). All camel bones are from the layers that, according to ceramics and other finds, date back to the end of the 3rd and the beginning of the 4th century Ad. In villa rustica in Vranj, one astragalus, one fourth

Fig. 2. — Ventral and dorsal views of the first cervical vertebra discovered in Viminacium amphitheatre. Scale in cm.

285



tarsal and the first phalanx were found in the lay-ers contemporary to those of Viminacium finds. According to the size and fusion of epiphyses, all specimens originate from adult animals.

In order to conclude whether they belong to C. bac-trianus or C. dromedarius, the measurements and morphology of our specimens were compared with contemporary camels which were studied in detail by C. Steiger (1990) in her thesis. We have also used metric data of ancient specimens from Roman and medieval sites in the region (Bartosiewicz&dirjec 2001; Bartosiewicz 1995) and from united Arab Emirates (uerpmann 1999). All measurements (Fig. 12) were taken following Von den driesch (1976) and Steiger (1990). As we included both morphology and metrics in species identification, specimens that could not be identified accord-ing to both criteria remained indeterminate. The context and dating of finds, historical sources and archaeological finds were considered in the expla-

Table 1. — Table of measurements of studied camel specimens. Measurements taken following Von den Driesch (1976). Abreviations: GL, greatest length/greatest length of lateral half; GLm, greatest length of medial half; GB, greatest breadth; GLF, greatest length from the Facies articularis cranialis to the Facies articularis caudalis; BFcr, breadth of the Facies articularis cranialis; BFcd, breadth of the Facies articularis caudalis; Bp, breadth of the proximal end; Dp, depth of the proximal end; SD, smallest breadth of the diaphysis; Bd, breadth of the distal end; BFd, breadth of the Facies articularis distalis; Dd, depth of the distal end; Dl, depth of the lateral half. Measurements taken following Steiger (1990): Lfavr, length of fossa alaris ventralis, BrFd, length of articular surface for oscarpi radiale; BuFd, length of facies articularis for os carpi ulnare; BFOm, breadth of articular facet for Os malleolare.

site

ske-

leta

l p

art

GL

GLm

GB

GLF

BFc

r

BFc

d

Lfav

r

Bp

Dp

SD

Bd

BFd

Dd

BrF

d

BuF

d

BFO

m

Dl

Viminacium, Amphitheatre

Atlas 145.2 126.6 110 94.5 92.6 16


Radius 104 88.2 32 40


Femur 128


Tibia 50.8 81.7 51.2 20.4

Viminacium, Amphitheatre Astragalus

80.4 74.9 54.3 44.6

Vranj, villa rustica

77.7 72.7 51.6 44.5


Os tarsale quartum

59


Ph I -posterior

100 38.2 30.6 20.6 34.9

Viminacium, Pirivoj

96.3 41.2 32 23.4


20.4

nation of the presence of camels in the Roman provinces in Serbia.

tAxonoMy And MoRPHoMEtRICS oF CAMEL BonES

Bactrian and dromedary camels differ in their adap-tations to different temperature conditions. drom-edaries which live in hot deserts, have shorter hair and generally longer limbs in contrast to Bactrians, adopted to colder climate, which have more mas-sive stature (Köller-Rollefson 1991). Consequently, differences in morphology and proportions allow a distinction between these two species in almost all of the postcranial bones (Steiger 1990). However, as archaeological material is usually fragmented, osteological differentiation between dromedaries and Bactrians is not a simple task. Identifying the species of ancient camel bones is further compli-



cated by possible appearance of hybrids, as camel hybridization has been practiced from the first centuries Ad until recently (uerpmann 1999).

Vertebrae and ribS

An atlas (Fig. 2), thoracic and lumbar vertebrae, and one rib were found together with a first anterior phalanx in the area of a room next to the western entrance of the amphitheatre. As they originate from the same context, we suppose, at least for the vertebrae, that they originate from a single animal.

The measurements of the atlas (table 1) were compared with the mean values of Bactrian camels and dromedaries from a comparative material studied by Steiger (1990) in Simpson’s (1941) ratio diagram (Fig. 3). Bactrian males were taken as the reference value and logarithms of their dimensions were placed on the ‘0’ line of the graph, and arithmetic differ-ences between them and the logarithms of female

AD

B

C

Fig. 4. — Camel specimens from Viminacium amphitheatre: A, Radius, posterior view; B, Femur, medial view; C, Tibia, anterior view, and villa rustica in Vranj; D, 4th tarsus, proximal view. Scale in cm.

GL0.05

0

-0.1

-0.15

-0.2

-0.25

-0.3

-0.05GB

GLF

BFcr

BFcd

Lfavr

C. bactrianus, female (Steiger 1990)

C. dromedarius, male (Steiger 1990)C. dromedarius, female (Steiger 1990)C. dromedarius x C. bactrianus, Viminacium

Fig. 3. — Simpson’s (1941) ratio diagram. This graph shows arithmetic differences between the reference values (logarithms of measurements of contemporary male bactrians) and logarithms of the specific measurements of female bactrians, male and female dromedaries and Vimiancium specimen. Abbreviations: GL, great-est length; GB, greatest breadth; GLF, greatest length from the Facies articularis cranialis to the Facies articularis caudalis; BFcr, breadth of the Facies articularis cranialis; BFcd, breadth of the Facies articularis caudalis; Lfavr, length of fossa alaris ventralis.

287



Bactrians, female and male dromedaries and Vimi-nacium specimen were shown. Larger dimensions are above the reference line, while smaller ones are below it. This graph illustrates that almost all the dimensions of theViminacium atlas fall within or are even greater than mean values of male Bactrians. However, proportions of Viminacium atlas cor-respond to dromedaries. The length of fossa alaris ventralis, one of the morphometric features that apparently separates the two camel species (Steiger 1990:14), is 16mm in our specimen, which is the mean value for dromedaries, while Bactrians have an aperture almost twice as big. Foramen vertebrale laterale is divided into two parts only on one side of our specimen. This foramen is divided into two or three parts in dromedaries, while in Bactrians there should be only one on each side. A Bactrian specimen studied by Steiger (1990) was an excep-tion, with two apertures on one side and one on the other, like in our specimen. our specimen also manifests mixed features in the morphol-ogy of foramen alare, which is developed on one side, while there is only an incisures on the other. The presence of foramen alare in our specimen is a feature typical of dromedaries, while Bactrians have only incisura instead. other morphological features mainly correspond to the morphology of

Fig. 5. — Anterior aspect of astragali from: A, VIminacium amphitheatre; B, Vranj. Scale in cm.

A B

one-humped camels. As this specimen is extremely robust, with the measurements that exceed those of Bactrians and proportions of dromedaries, and since it manifests morphologically mixed features between those of dromedaries and Bactrians, we have determined it to be a hybrid, Camelus drom-edarius x Camelus bactrianus.other vertebrae: thoracic and lumbar, along with the rib do not have enough characteristics for reli-able species identification.

radiuS and ulna

A distal epiphysis with preserved half shaft of right fused radius and ulna (Fig. 4A) were found within the inner walls of the eastern part of Viminacium amphitheatre. Measurements of the radius – the breadth of distal epiphysis (104mm) and the breadth of facies articularis (88.2mm) both correspond to a large male two-humped camel and exceed the measurements of contemporary dromedaries studied by Steiger (1990: 94). on the other hand, the morphology of the distal shaft and articulation suggest a one-humped camel: the length of facies articularis for os carpi ulnare (40 mm) is larger than the length of articular surface for os carpi radiale (32 mm). Since the morphology and proportions correspond to a one-humped camel and measure-



ments correspond to a two-humped camel, these bones are ascribed to a hybrid C. dromedarius x C. bactrianus.

Femur A fragment of a distal shaft and epiphysis of a left femur (Fig.4B) was found in the southern part of the arena of Viminacium amphitheatre. The lateral part of the femur distal diaphysis has a distinguishing groove of dromedaries, while this groove is missing in Bactrians (Steiger 1990: 49). The lateral part of the distal shaft in our specimen is broken, therefore, identification of the species according to this morphological feature is not possible. The breadth (128 mm) of the distal epiphysis of our specimen exceeds measurements of dromedaries and corresponds to Bactrians (Steiger 1990: 97). Since metrics corresponds to a bactrian, but neither metrics

nor morphology exclude a hybrid, this specimen is identified as Camelus sp.

tibia

A fragment of a right tibia (Fig. 4C) with a pre-served distal end and almost half shaft was found in the area to the south of the Western entrance of Viminacium amphitheatre. The only morphological distinction in distal tibia pointed out by Steiger (1990: 54) is articular facet for Os malleolare, which is wider in Bactrians than in dromedaries. The breadth of the mentioned articular facet in our specimen (20.4 mm) exceeds the measurements of recent dromedaries and corresponds to Bac-trians studied by Steiger (1990: 99). The breadth of the distal epiphysis (81.7 mm) corresponds to male dromedaries and female Bactrians, while the depth (51.2 mm) of distal epiphysis exceeds the measurements of dromedaries and falls within the mean values of Bactrians. According to both mor-phology and metrics this tibia probably belonged to C. bactrianus.

aStragali

A complete right astragalus was discovered in the area south of the West entrance of Viminacium amphitheatre, while a complete left astragalus was found in the villa rustica in Vranj (Fig 5; table 1).

As the lateral part of trochlea tali extends fur-ther the proximal in dromedaries (Steiger 1990: 58), the ratio between lateral and medial astragali length is different in the two camel species. We have compared the lateral and medial length of astragali from Viminacium and Vranj with con-temporary specimens (Steiger 1990) and also with the closest eastern dromedaries and hybrids from the ancient site of Mleiha (uerpmann 1999) (Fig. 6). The proportions of both our specimens exclude dromedaries and correspond to Bactri-ans or hybrids. Vranj specimen measurements fall within the mean Bactrian values, while the measurements of Viminacium astragalus fall within the largest Bactrians and hybrids from Mleiha. According to morphological features that distinguish one- from two-humped camel (Steiger 1990), such as the width of the trochlea tali distalis, both Viminacium and Vranj astragali

C. bactrianus (Steiger 1990)

C. dromedarius (Steiger 1990)

C. dromedarius, Mleiha (Uerpmann 1999)

C. dromedarius, Szekszárd-Palánk (16-17th A.D.) (Bartosiewicz1996)

C. bactrianus, Vranj

C. bactrianus, ViminaciumGLm

78

76

74

72

70

68

66

64

62

6065 70 75 80 85

GLl

C. dromedarius x C. bactrianus, Mleiha (Uerpmann 1999)

Fig. 6. — The ratio between the greatest lateral length (GLl) and the greatest medial length (GLm) of astragali of contemporary and archaeological specimens.

289



Fig. 8. — The ratio between the greatest length (GL) and proximal epiphysis breadth (Bp) of the first posterior phalanges of archaeo-logical and contemporary specimens.

Fig. 7. — Cranial view of the 1st posterior phalanges from: A, Viminacium amphitheatre; B, Viminacium-Pirivoj; C, Vranj. Scale in cm.

A B C

C. dromedarius, Mleiha (Uerpmann 1999)C. dromedarius (Steiger 1990)C. bactrianus (Steiger 1990)

C. bactrianus, Viminacium-PirivojC. dromedarius x C. bactrianus,Viminacium-Amphitheatre

GL

Bp

44

42

40

38

36

34

32

75 80 85 90 95 100 105

west entrance. The third first posterior phalanx (Fig. 7C), which lacks distal epiphysis, was found at villa rustica in Vranj.

correspond to Bactrian camels. As both morphol-ogy and metric correspond to Bactrians, we have identified both astragali as C. bactrianus.

4th tarSuS An almost complete left fourth tarsus (Fig. 4d) was discovered at villa rustica in Vranj. The fourth tarsus differs in dromedaries and Bactrians by the morphology of articular facet, as it is established in modern camels (Steiger 1990: 64). The articular facet for the forth metatarsus is strongly angled in dromedaries, while this facet in Bactrians is wide and flat. According to this feature, our specimen corresponds to a Bactrian. Its size is concordant to mean values for Bactrians, but does not exclude a large male dromedary or a hybrid. Therefore, we identify this specimen as Camelus sp.

PhalangeS

In a 4th century trench in the dump of the East-ern necropolis of Viminacium, in the locality of Pirivoj, one first posterior phalanx (Fig. 7B)was found. Another first posterior phalanx (Fig. 7A)was found together with the aforementioned verte-brae and ribs in the room next to the amphitheatre



The first phalanges of C. bactrianus and C. drom-edarius differ in both metrics and morphology. drom-edaries have slimmer phalanges, which are stocky in Bactrians. The proximal epiphysis in two-humped camel tends to be wider, while in dromedaries it is more narrow and deep. Morphologically, the only distinguishing feature is the dorsal outline of the proximal articulation which is oval in Bactrians and has a beveled abaxial corner in dromedaries (Steiger 1990: 73). Specimens from Pirivoj and Vranj have

oval edges, while in the specimen from the amphi-theatre the abaxial edge is distinctly bevelled. The greatest length and proximal breadth of Viminacium phalanges, modern camels (Steiger 1990) and two dromedaries from Mleiha cemetery (uerpmann 1999) are plotted on the scatter diagram (Fig. 8).The phalanx from Pirivoj falls within the range of Bactrians. The amphitheatre specimen has the proportions of dromedaries, but its length exceeds the lengths of phalanges of both species. According to both morphology and metrics, Viminacium-Pirivoj specimen is determined as C.bactrianus. The Am-phitheatre specimen morphologically corresponds to dromedaries, but its measurements exceed both species, so we have determined this specimen was a hybrid, C. dromedarius x C. bactrianus. Vranj specimen remained indeterminate.

tAPHonoMy oF CAMEL REMAInS

Long bones, radius and ulna, femur and tibia are fragmented, while short bones, ribs and vertebrae are almost complete. This is in accordance with meat and bone marrow exploitation. Butchering marks were found on the atlas, radius/ulna and tibia. There is an artificially made puncture (di-ameter c.15mm) on the ventral plate of the atlas wing, which was probably made by a leaf-shaped spear. Three slice marks were located at the pos-terior side of the radius mid shaft. on the tibia, there are disarticulation cut-marks on the cranio-medial aspect of the distal shaft and chop marks along the breakage. Those butchering marks suggest that camel meat was consumed to some extent. Good preservation of bones indicates that they were not exposed to external conditions before burial. Gnawing marks were noticed on the distal end of the phalanx from the Eastern necropolis.

dISCuSSIon: WHAt WERE tHE CAMELS uSEd FoR?

camelS in military uSe

dromedary camels and dromedarii, camel riders, were included in Roman units in Egypt and Syria

Fig. 9. — Detail of frieze of Constantine arch depicting Constantine riding his chariot, with the army departing from Milan to attack Maxentius. Besides army and horses, the procession included a camel. Although, it is poorly presented, with the neck which is too short and an almost invisible hump, the shape of the head, its size and the load on its back indicate that it is a camel. (The photo was taken from web page http://www.rome101.com/Topics/ArchConstantine/ (1.11.2011.)

Fig. 10. — Illustration of different ways of utilization of camels, drawn by B. Hawkins and published by T. Varty c. 1850, in Graphic Illustrations of Animals and Their Utility to Man.

291



in the 1st and 2nd centuries Ad and they were rein-troduced during the time of Constantine (toynbee 1996) (Fig. 9). The Roman army used camels for transporting heavy objects: large supplies of corn, road building equipment, luggage and military equipment, but also letters (davies 1967: 117). As other beasts of burden, camels were obtained from civilians and trained by the army. A specified standard regarding animals stature and health was required and they were subjected to veterinary examination before inclusion in the units (davies 1989: 153-156). Except as pack animals, the Ro-man Army also used dromedaries to frighten horses (Bohec 2000: 27). The atlas with an impact mark probably made by a leaf shaped arrow could indicate that the camel died in battle, so the military use of the camels in Viminacium is also possible. As it was described before, that specimen originates from a large individual, probably a hybrid, certainly of great quality, which is in accordance with Roman military requirements.

camelS aS long-diStance tranSPortation animalS

As animals that carried loads, either for armies or in caravan trade, camels were mentioned by ancient writers (eg: Plin. nat. 8.26; Strabo xVII, 45) and portrayed in Roman works of art (toyn-bee 1996). Sirmium and Viminacium were both located on main roads that connected Aquilea and the northern parts of the Empire with the Black sea regions, and also with Greece, so that caravans from all over the Empire must have passed this way. Archaeological finds confirm that during the 3rd and the 4th century Ad in Moesia Prima and Pannonia Secunda intensive trade import was going on of cereals, vine, oil and other products from the Black sea region, and to a minor extent from the north Africa and Asia Minor. Great quantities of amphorae that carried wine and olive oil from the Black sea region were imported between the 3rd and the 6th century Ad to this region. A small number of amphorae came from north-African region, Egypt and tu-nisia, and it is believed that, except for wine and olive oil, they also carried fish, honey, ethereal oil and perfumes. Black Sea imports were prob-

ably shipped by waterways, whilst north African goods must have traveled along different and also terrestrial routes (Bjelajac 1996).

Long distance trade was common to the Roman world and it included, aside from luxury objects, low cost products, such as wine, oil, fish, etc. From the time of Augustus, the import of silk from China was intensified (Thorley 2011:71), and camels certainly carried a big portion of the goods from faraway China, along the Silk Road.

camelS in roman gameS

From historical sources it is known that Emperor Claudius (dio Lx 7 3) organized camel fights and that Emperor nero (Suetonius nero III) introduced camel races in Circus Maximus (toynbee 1996). Camel remains from Viminacium amphitheatre are from the period when this edifice had already lost its function, since they are from the layers that buried the building. Since there are no camel remains from the 2nd and 3rd century Ad in Viminacium, we cannot presume that camels participated in pub-lic shows in Viminaciumamphitheatre. Elsewhere in the Roman world, camel bones were found in amphitheatres: in Flavian Amphitheatre in Rome (de Grossi Mazzorin et al. 2005), in the Roman amphitheatre of Cartago nova in Spain (Morales Muñiz et al. 1995) and in the Roman amphithea-tre of Serdica in Bulgaria (Velichkov 2009: 125).

camel meat and Secondary ProductS conSumPtion

Galen (AF 6.666) wrote that camel was eaten only by people who were ‘mentally and physically cam-el-like’, while Aristotle (Hist. Anim. B. III. c. 20) wrote about delicate taste of camel milk and meat (dalby 2003: 71). Pliny wrote (Plin. nat.28.33) that camel products, such as brain, tail and dung were used in medical and beauty treatments. Meat, milk, fat, hair, wool, leather, dung and other products of dromedaries (Köller-Rollefson 1991) and Bactrians (Potts 2004) have been in use ever since ancient times. Well-fed camels can produce large quantities of meat and fat. Wool of Bactrian camel is valued because of great quality of fibre (Peters & driesch 1997). dromedary wool is used for making blan-kets and tents, while sandals, whips and saddles are



made of leather. Both dromedaries and Bactrians are also used for traction (Köller-Rollefson 1991; Potts 2004). Fragmentation and butchering marks on long bones from Viminacium suggest that camel meat was consumed. In comparison with other domestic animals which were kept for meat, milk, wool and traction, camels were in much smaller number. All the camel bones discovered come from adult animals, so we can hypothesize that camels were butchered when they were too old2 to serve as pack or riding animals, which was probably their primary function.

ConCLuSIon

Among 17 camel bones from Viminacium and Vranj (table 1), we have identified 6 bactrian remains, 3 hybrids, while the remaining bones were identi-fied as Camelus sp., since they did not have enough morphologically distinctive features.

Because of its great size and mixed features, we suggest that the first cervical vertebra, phalanx and radius/ulna found in Viminacium amphitheatre originate from hybrid individuals. In the horse and camel graveyard at Mleiha in united Arab Emir-

2. Two-humped camel usually live 35-40 years, they are put to work at the age of four and they work for about 20-25 years (Potts 2004: 147, after Epstein 1969).

ates, which dates back to the 1st-2nd century Ad, according to their robustness and morphologically mixed features between dromedaries and Bactrians, three large camel individuals were determined as hybrids between these species. As they had the central position in the graveyard, the author sug-gested that they were status animals (uerpmann 1999). In the site of Pella, decapolis in Jordan, the skeletal remains of camels which died in the earthquake in 747 Ad, were also determined as hybrids (uerpmann 1999, after Köhler-Rollefson 1989). Additionally, the first anterior phalanx from Roman layers in troy, turkey, is also designated as a hybrid specimen (uerpmann 1999: 113). Although there are no reported archaeozoological evidences of crossbreeds between two camel species in Roman period sites in other parts of Europe, L. Bartosie-wicz (Bartosiewicz & dirjec 2001:284) wrote that large camel specimens from Central Europe could either be the remains of castrated animals selected for the army, or of hybrids. Hybrids are larger and stronger, but look more like dromedaries as they have only one hump, which is usually asymmetri-cal or flattened at the top. They are about 2.15 m high at the shoulder or 2.32 m at the hump. They have long legs and their height is usually greater than their length, their average weight is 650 kg, but they can reach 900-950 kg in weight (Potts

Table 2. — Table of studied camel specimens.

site skeletal part taxon

Viminacium, Amphitheatre Atlas Camelus dromedarius x Camelus bactrianus fig. 2Viminacium, Amphitheatre Vertebra thoracica Camelus sp. /Viminacium, Amphitheatre Vertebra thoracica Camelus sp. /Viminacium, Amphitheatre Vertebra thoracica Camelus sp. /Viminacium, Amphitheatre Vertebra lumbalis Camelus sp. /Viminacium, Amphitheatre Vertebra lumbalis Camelus sp. /Viminacium, Amphitheatre Vertebra lumbalis Camelus sp. /Viminacium, Amphitheatre Costa Camelus sp. /Viminacium, Amphitheatre Radius and ulna Camelus dromedarius x Camelus bactrianus fig. 4AViminacium, Amphitheatre Femur Camelus bactrianus fig. 4BViminacium, Amphitheatre Tibia Camelus bactrianus fig. 4CViminacium, Amphitheatre Astragalus Camelus bactrianus fig. 5AViminacium, Amphitheatre 1st phalanx Camelus dromedarius x Camelus bactrianus fig. 7AViminacium, Pirivoj 1st phalanx Camelus bactrianus fig. 7BVranj, villa rustica Astragalus Camelus bactrianus fig. 5BVranj, villa rustica 1st phalanx Camelus sp. fig. 7CVranj, villa rustica Os tarsale quartum Camelus bactrianus fig. 4D

293



2004, after Koplakow 1935). In ethnographic sources from Central Asia from the 17th century, it is noted that hybrids could carry 500 kg, which is twice as much as dromedaries (Potts 2004 after tapper 1985). So, hybrids have greater strength and load-bearing abilities than both species.

taxonomic identification of Viminacium and Vranj camels is important because camel species originate from different parts of the world: Bac-trians from Central Asia and dromedaries from north Africa and Western Asia. Although utiliza-tion of both camel species is similar, there is a slight difference. Bactrians are mainly used as pack and draught animals (Potts 2004:1950), while drom-edaries are used for riding, as beasts of burden and for traction (Köller-Rollefson 1991). The presence of two-humped camels in Roman sites is probably related to civilian movements, i.e. to long distance caravans from Central Asia (Bartosiewicz&dirjec 2001), while one-humped camel remains probably refer, aside to civilian, to military usage. Accordingly, their arrival in Roman provinces of Moesia Prima and Pannonia Secunda has a different connota-tion. Ethnographic examples showed that hybrid camels were bred in regions such as turkmenistan, Afghanistan and Iran, where both dromedaries and Bactrians co-existed (Köller-Rollefson 1991). Since there are not enough archaeozoological data on camel hybridization from ancient times, the origin of hybrids identified in Serbia remains uncertain.

The appearance of those strong and capable ani-mals in Roman provinces, in caravans, that carried trade goods or in Roman army units and convoys is more than reasonable. As hybrids were well adapted to colder climate and muddy terrain (Potts 2004, after tapper 1985), they could certainly withstand European climate better than dromedaries. The habitat of Bactrian camels includes arid continen-tal areas, with large differences in temperature, so they could put up with high and low temperatures (Grzimek 2003: 316) and certainly could manage quite well in the European continental climate.

Although there are few camel bones in regard to other archaeozoological material from Roman period sites in Serbia, we presume that the camel was not a very rare animal in the region. It should be mentioned that while finishing this paper, new

camel bones were unearthed in Viminacium and that four new camel bones were also discovered in Southern Serbia, in davidovac near Vranje, and also in Pirot, in Eastern Serbia. Just a small part of the Roman period fauna from Serbia has been pro-cessed until now (Blažić 1993; Blažić 1995; Blažić 2006; Bökönyi 1976; Gilić 1994; dimitrijevic & Medovic 2007; nedeljković 1997; nedeljković 2009; Vuković 2010) and the presence of camel bones on four sites already confirms the assumption that camels were present in the provinces in today’s Serbia not in such a small number.

to conclude, camels came here in caravans car-rying trade goods or with Roman army units and convoys as riding or pack animals in the beginning of the 4th century Ad. Along main roads in the Ro-man provinces of Moesia and Pannonia, along with horses and mules, but in smaller number, camels were certainly part of not a very unusual landscape.

AcknowledgmentsWe are grateful to Miomir Korać, Phd, director of the Project of Viminacium for allowing us to work on animal bone material from the site. We are also very grateful to archaeologists Snežana nikolić and Ivan Bogdanović, participants in the Project of Viminacium, for sharing with us recent results on Viminacium amphitheatre. Special thanks go to Roel Lauwerier, who was very kind to send us his unpublished Phd thesis on animal remains from Sirmium. Thanks also go to Prof. dr. Vesna dimitrijević, for reviewing the paper and giving us her valuable comments. Finally, special thanks go to Mrs. Rajka Marinković, Professor of English, for linguistic corrections.

This paper is the result of the Project „Bioar-chaeology of Ancient Europe – humans, animals and plants in the prehistory of Serbia’ (III 47001) funded by the Ministry of Education and Science of the Republic of Serbia.

REFEREnCES

aPPlebaum S. 2002. — Animal husbandry, in Wacher J. S. (ed.), The Roman World II. Routledge, London: 504-526.



bartoSieWicz l. 1995. — Camel Remains from Hun-gary, in buitenhuiS h. & uerPmann h.-P. (eds), Archaeozoology of the Near East II. Backhuys Publish-ers. Leiden: 119-125.

bartoSieWicz l. 1996. — Camels in antiquity: the Hungarian connection. Antiquity 70: 447-453.

bartoSieWicz l. 1999. — Recent developments in archaeozoological research in Slovenia. Arheološki vestnik 50: 311-322.

bartoSieWicz l. & dirjec j. 2001. — Camels in an-tiquity: Roman Period finds from Slovenia. Antiquity 75: 279-85.

beech m. 2007. —The Large Mammal and Reptile Bones, in Poulter a. g. (ed.), Nicopolis ad Istrum. A Late Roman and Early Byzantine City. The Finds and the Biological Remains. oxbow Books on behalf of the Society of Antiquaries of London, London: 154-198.

benecke n. 1994. — Der Mensch und seine Haust-iere. Die Geschichte einer jahrtausendalten Beziehung. Parkland-Verlag, Stuttgart.

bjelajac l. 1996. — Amphorae of the Danubian Basin in Upper Moesia. Archaeological Institute Belgrade, Belgrade.

blazic S. 2006. — Fauna arheoloških lokaliteta u Vojvo-dini. MuzejVojvodine, novi Sad.

blazic S. 1993. — ostaci životinja sa lokalitetaVranj 1991. Rad Muzeja Vojvodine 35: 71-78.

blazic S. 1995. — ostaci životinjskih vrsta sa lokaliteta na trasi auto-puta kroz Srem, in VaPa z. (ed.), Arheološka istraživanja duž auto-puta kroz Srem. Pokrajinski zavod za zaštitu spomenika culture, novi Sad: 331-346.

bohec y. l. 2000. — The imperial Roman army. Rout-ledge, London.

bökönyi S. 1989. — Camel Sacrifice in Roman Intercisa. Acta Archaeologica Scientiarium Hungarica 41: 399-404.

bökönyi S. 1974. — History of Domestic Mammals in Central and Eastern Europe. Akadémiai Kiadó, Budapest.

bökönyi S. 1976. — ostaci životinjskih kostiju u dumb-ovu – lokalitetu u Fruškoj gori iz rimskog carskog perioda. Preliminarni izveštaj. Građa za proučavanje spomenika kulture Vojvodine 6-7: 49-51.

clutton-brock j. 1987. — A natural history of domesti-cated animals. Cambridge university Press, Cambridge.

dalby a. 2003. — Food in the ancient world from A to Z. Routledge, London.

daVieS r.W. 1967. — “Ratio” and “opinio” in Roman Military documents. Historia 16: 115-118.

daVieS r.W. 1989. — Service in the Roman Army. Ed-inburgh university Press, Edinburgh.

de Grossi mazzorin j. 2006. — Cammelli nell ’ an-tichità: le presenze in Italia, in Sala b. & tecchiati u. (eds), Archaeozoological studies in honour of Alfredo Riedel. ufficio Beni archeologici, Bolzano: 231-242.

de groSSi mazzorin j. 2011. — Presenze di cammelli nell’Antichità in Italia e in Europa: aggiornamenti,

in VolPe g., buglione a. & de Venuto g. (eds.), Vie degli animali, vie degli uomini. Transumanza e altri spostamenti di animali nell’Europa tardoantica e medievale. Edipuglia s.r.l., Bari: 91-106.

de groSSi mazzorin j., minniti c. & rea r. 2005. — de ossibus in anphitheatro Flavio effossis: 110 anni dopo i rinvenimenti di Francesco Luzj, in malerba g. & ViSentini P. (eds), Atti del 4° Convegno Nazionale di Archeozoologia, Pordenone, 13-15 novembre 2003. Quaderni del Museo Archeologico del Friuli Occiden-tale 6. Comune di Pordenone, Museo Archeologico, Pordenone: 337-348.

dimitrijeVic V. & medoVic a. 2007. — Animal and plant remains in a tomb in test-pit 1/05 outside the fortified imperial palace Felix Romuliana. Starinar 59: 315-323.

Von den drieSch a. 1976. —- A Guide to the Mea-surement of Animal Bones from Archaeological Sites. Harvard university, Harvard.

gilic d. 1994. — ostaci konja (Equus caballus L.) sa arheološkog lokaliteta Hipodrom u Sirmijumu. Rad Muzeja Vojvodine 36 : 81-94.

grzimek b. 2003. — Camels, guanacos, llamas, alpacas and vicuñas. Grzimek’s Animal Life Encyclopedia. 2nd edition. Volumes 12–16 Mammals I–V :313-323.

korac m. & goluboVic S. 2009. — VIMINACIUM. Više grobalja 2: Archaeological Institute Belgrade, Belgrade.

köller-rolleFSon i. 1991. — Camelus dromedarius. Mammalian Species 375: 1-8.

köller-rolleFSon i. 1993. — Camels and Camel Pastoralism in Arabia. Biblical Archaeologist 56(4): 180-188.

köller-rolleFSon i. 1996. — The one-humped camel in Asia: origin utilization and mechanisms of di-spearsal, in harriS d. r. (ed.), The origins and spread of agriculture and pastoralism in Eurasia. Smithsonian, Washington : 282-294.

lauWerier r. c. g. m. 1978. — Dieren in Sirmium. unpublished Phd Thesis. Biologisch-Archaeologisch Instituut, university of Groningen, Groningen.

milošeVic P. 2001. — Arheologija i istorija Sirmijuma. Matica Srpska, novi Sad. Morales Muñiz A., RiquelmE J.A. & Liesau von Lettow-Vorbeck C. 1995. - drom-edaries in Antiquity : Iberia and beyond. Antiquity 69: 368-375.

nedeljkoVic d. 1997. — ostaci životinjskih kostiju sa lokaliteta 80 Sirmijuma (1996.g.). Preliminarni izveštaj. Zbornik Muzeja Srema 3: 37-45.

nedeljkoVic d. 2009. — Pregled arheozooloških istraživanja Sirmijuma – lokalitet 85 (2002-2005). Zbornik Muzeja Srema 8 : 7-42.

nikolic S. & bogdanoVic i. 2012. — Istraživanja viminacijumskog amfiteatra u toku 2011. godine, in bikic V., goluboVic S. & antonoVic d. (eds), Ar-heologija u Srbiji: Projekti Arheološkog instituta u 2011.

295



SimPSon g. 1941. — Large Pleistocene Felines of north America. American Museum Novitates 1136: 1-27.

Steiger c. 1990. — Vergleichend morphologische Unter-suchungen an Einzelknochen des postkranialen Skeletts der Altweltkamele. Phd Thesis. Ludwig-Maximilians-universität, Munich.

thorley j. 2011. — The Silk trade Between China and the Roman Empire at its Height “Circa” A. d. 90-130. Greece & Rome 18(1): 71-80.

toynbee j. m. c. 1996. — Animals in Roman life and art. Johns Hopkins university Press, Baltimore.

uerPmann h.-P. 1999. — Camel and horse skeletons from protohistoric graves at Mleiha in the Emirate of Sharjah (u.A.E.). Arabian Archaeology and Epigraphy 10(1): 102-118.

VelichkoV z. 2009. — The amphitheatre of Serdica (City of Sofia Bulgaria), in Wilmott t. (ed.), Roman Amphitheatres and Spectacula: a 21st Century Perspec-tive. BAR International Series 1946. Archaeopress, oxford: 119-125.

VukoVic S. 2010. — ostaci životinja sa jugozapadnog dela lokacije Pirivoj (Viminacijum). Arheologija i prirodne nauke 5: 57-82.

VukoVic S. & bogdanoVic i. 2013. — A Camel Skel-eton from the Viminacium Amphitheatre. Starinar 63: 251-267.

zeder m. a., emShWiller e., Smith b. d. & brad-leyd. g. 2006. — documenting domestication: the intersection of genetics and archaeology. Trends in Genetics 22(3): 139-55.

zeuner F. e. 1967. — Geschichte der Haustiere. Bayer-ischer Landwirtschaftsverlag, München - Basel - Wien.

Submitted on January 2013; accepted on April 2013.

godini. Arheološki institut, Beograd: 42-45. olSen S. j. 1988. — The Camel in Ancient China and

an osteology of the Camel. Proceedings of the Academy of Natural Sciences of Philadelphia 140(1): 18-58.

PeterS j. & Von den drieSch a. 1997. — The two-humped camel (Camelus bactrianus): new light on its distribution management and medical treatment in the past. Journal of Zoology 242: 651-679.

Pigière F. & henrotay d. 2012. — Camels in the northern provinces of the Roman Empire. Journal of Archaeological Science 39 (5): 1-9.

PottS d. t. 2003. — Bactrian Camels and Bactrian-dromedary Hybrids. Silkroad 3(1). Available at: http://www.silk-road.com/newsletter/vol3num1/7_bactrian.php.

PottS d. t. 2004. — Camel Hybridization and the Role of Camelus bactrianus in the Ancient near East. Journal of the Economic and Social History of the Ori-ent 47(2): 143-165.

riedel a. 1999. — Kamelfunde der frühen römischen Kaiserzeit aus Mauerbach (niederösterreich, Bezirk Wien-umgebung). Annalen des Naturhistorischen Museums Wien 100: 81-92.

rušeVljan V. d. 2004. — Vila rustika u Hrtkovcima. Sistematsko-zaštitno iskopavanje. Zbornik Matice srpske za klasične studije 6: 171-173.

rušeVljan V. d. 2005. — Sistematsko-zaštitna iskopa-vanja vile rustike u Hrtkovcima 2004. godine. Glasnik Srpskog arheološkog društva 22(21): 239-249.

Schramm z. 1975. — Zwierzece szczatki kostne (The Animal Bones), in Parnicki-Pudelko S. (ed.), No-vae – Sektor Zachodni 1972. uniwersytet im. Adama Mickiewicza w Poznaniu, Poznan: 216-241.

camels from roman imperial sites in serbia

Documents